12th Annual Conference of the International Speech Communication Association

Florence, Italy
August 27-31. 2011

Factored MLLR Adaptation for Singing Voice Generation

June Sig Sung, Doo Hwa Hong, Shin Jae Kang, Nam Soo Kim

Seoul National University, Korea

In our previous study, we proposed factored MLLR (FMLLR) where each MLLR parameter is defined as a function of a control vector. We presented a method to train the FMLLR parameters based on a general framework of the expectation-maximization (EM) algorithm. In this paper, we extend the FMLLR structure from diagonal to unrestricted full matrix with a sophisticated algorithm for the training of relevant parameters. In the experiments on artificial generation of singing voice, we evaluate the performance of the FMLLR technique with two matrix structures and also compare with other approaches to parameter adaptation in HMM-based speech synthesis.

Full Paper

Bibliographic reference.  Sung, June Sig / Hong, Doo Hwa / Kang, Shin Jae / Kim, Nam Soo (2011): "Factored MLLR adaptation for singing voice generation", In INTERSPEECH-2011, 2789-2792.